Pulmonary artery catheterization (PAC) is often performed to obtain pulmonary artery wedge pressure (WP) measurements. The WP is frequently used as a therapeutic endpoint, as, for example, during fluid resuscitation of shock or for diuretic therapy in treating pulmonary edema. An alternative approach might be to use direct measurements of the extravascular water content of the lung (EVLW) as a therapeutic endpoint. We have already shown in a previous study (Am Rev Respir Dis 1987; 136:662-668), that a therapeutic protocol incorporating EVLW measurements can be carried out safely in critically-ill patients. This study also suggested possible beneficial effects on patient morbidity and mortality of following a protocol guided by EVLW measurements. To provide further support for these potentially important findings, we plan to compare an EVLW based protocol with the standard approach based on wedge pressure measurements, in a randomized prospective evaluation. This study will answer the following specific questions: 1) is the accumulation or resolution of pulmonary edema affected by patient management based on EVLW rather than WP measurements? 2) is patient outcome affected by patient management based on EVLW measurements?, and 3) is EVLW an important prognostic index in patients with pulmonary edema? We plan to enroll 80-100 patients. The data will be analyzed to determine whether treatment groups (EVLW vs WP) are different with respect to: 1) first and last EVLW measurement (normalized for differences in study duration) 2) time required for mechanical ventilatory support, time in the ICU, and both ICU and hospital mortality. 3) mortality as a function of initial EVLW measurement, or as a function of change in the EVLW measurement after treatment. From such data, we will determine whether the resolution of or outcome from pulmonary edema can be improved-by using a treatment protocol based on EVLW measurements, and whether such measurements can help define the prognosis of patients with severe pulmonary edema.